A tree regression modeling for assessment of effective factors on failure of arteriovenous fistulas among hemodialysis patients attended in Hasheminejad Kidney Center. RJMS 2018; 25 (9) :8-16
URL:
http://rjms.iums.ac.ir/article-1-4980-en.html
Abstract: (2911 Views)
Background: Hemodialysis is the sole way for management of patients with End Stage Renal Disorders (ESRD). Nowadays, inserting arteriovenous fistula is one of the most common methods for effective hemodialysis. Several factors have been reported for failure of arteriovenous fistulas such as diabetes, biochemical factors, low hemoglobin and increased level of blood calcium and phosphorus. Decision tree regression modeling is more useful among clinical studies. Tree regression modeling was used for present study for assessment of causes of failure in arteriovenous fistula among hemodialysis patients who referred to Hasheminejad Kidney Center.
Methods: The material of this historical Cohort study was gathered from records of hemodialysis patients with active and failed arteriovenous fistulas. Bivariate analysis and logistic regression analysis was performed and tree regression model created.
Results: There is significant association between arteriovenous fistula success and diabetes, hypertension among hemodialysis patients (p<0.001). Although we found significant association between serum level of hemoglobin and arteriovenous fistula outcome in bivariate analysis, but regression analysis showed only age, diabetes and hypertension could be looked upon as independent predictors of arteriovenous fistula success.
Conclusion: Hypertension and diabetes mellitus have significant roles in the outcome of AVF failure. Regarding the effect of “age” variable, we recommend further studies about AVF maturation, based on different age groups of hemodialysis patientsthe learning environment in order to increase students' motivation and they can achieve the success.
Type of Study:
Research |
Subject:
General Surgery